From ad92ca25c0ecbe155ddbb7153fded36e701fbd2b Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?H=C3=A9l=C3=A8ne=20Jonin?= <helene.jonin@gmail.com>
Date: Fri, 2 Aug 2024 14:02:02 +0000
Subject: [PATCH] Add cff to bibtex converter
---
swh/indexer/bibtex.py | 44 ++++++--
swh/indexer/metadata_dictionary/cff.py | 4 +-
.../tests/metadata_dictionary/test_cff.py | 1 +
swh/indexer/tests/test_bibtex.py | 101 +++++++++++++++++-
4 files changed, 141 insertions(+), 9 deletions(-)
diff --git a/swh/indexer/bibtex.py b/swh/indexer/bibtex.py
index dba65dcb..a232ee49 100644
--- a/swh/indexer/bibtex.py
+++ b/swh/indexer/bibtex.py
@@ -6,14 +6,16 @@
import collections
import json
import sys
-from typing import Any, Dict, List
+from typing import Any, Dict, List, Optional
import uuid
from pybtex.database import Entry, Person
import rdflib
from swh.indexer.codemeta import compact, expand
+from swh.indexer.metadata_dictionary.cff import CffMapping
from swh.indexer.namespaces import RDF, SCHEMA, SPDX_LICENSES
+from swh.model.swhids import ObjectType, QualifiedSWHID
TMP_ROOT_URI_PREFIX = "https://www.softwareheritage.org/schema/2022/indexer/tmp-node/"
"""IRI used for `skolemization <https://www.w3.org/TR/rdf11-concepts/#section-skolemization>`_;
@@ -21,7 +23,9 @@ it is not used outside :func:`codemeta_to_bibtex`.
"""
-def codemeta_to_bibtex(doc: Dict[str, Any]) -> str:
+def codemeta_to_bibtex(
+ doc: Dict[str, Any], swhid: Optional[QualifiedSWHID] = None
+) -> str:
doc = compact(doc, False)
identifiers = []
@@ -54,15 +58,23 @@ def codemeta_to_bibtex(doc: Dict[str, Any]) -> str:
fields: Dict[str, Any] = {}
def add_person(persons: List[Person], person_id: rdflib.term.Node) -> None:
+ person = Person()
for _, _, name in g.triples((person_id, SCHEMA.name, None)):
if (person_id, RDF.type, SCHEMA.Organization) in g:
# prevent interpreting the name as "Firstname Lastname" and reformatting
# it to "Lastname, Firstname"
- person = Person(last=name)
+ person.last_names.append(name)
else:
person = Person(name)
- if person not in persons:
- persons.append(person)
+
+ for _, _, given_name in g.triples((person_id, SCHEMA.givenName, None)):
+ person.first_names.append(given_name)
+
+ for _, _, family_name in g.triples((person_id, SCHEMA.familyName, None)):
+ person.last_names.append(family_name)
+
+ if str(person) and person not in persons:
+ persons.append(person)
def add_affiliations(person: rdflib.term.Node) -> None:
for _, _, organization in g.triples((person, SCHEMA.affiliation, None)):
@@ -160,7 +172,20 @@ def codemeta_to_bibtex(doc: Dict[str, Any]) -> str:
for _, _, version in g.triples((id_, SCHEMA.version, None)):
fields["version"] = version
- entry_type = "softwareversion" if "version" in fields else "software"
+ # entry_type
+ if swhid:
+ fields["swhid"] = str(swhid)
+ if swhid.object_type == ObjectType.SNAPSHOT:
+ entry_type = "software"
+ elif swhid.object_type == ObjectType.CONTENT:
+ entry_type = "codefragment"
+ else:
+ entry_type = "softwareversion"
+ elif "version" in fields:
+ entry_type = "softwareversion"
+ else:
+ entry_type = "software"
+
entry = Entry(
entry_type,
persons=persons,
@@ -171,6 +196,13 @@ def codemeta_to_bibtex(doc: Dict[str, Any]) -> str:
return entry.to_string(bib_format="bibtex")
+def cff_to_bibtex(content: str, swhid: Optional[QualifiedSWHID] = None) -> str:
+ codemeta = CffMapping().translate(raw_content=content.encode("utf-8"))
+ if codemeta is None:
+ codemeta = {}
+ return codemeta_to_bibtex(codemeta, swhid)
+
+
if __name__ == "__main__":
for filename in sys.argv[1:]:
if filename == "-":
diff --git a/swh/indexer/metadata_dictionary/cff.py b/swh/indexer/metadata_dictionary/cff.py
index 0d730e88..93d32b21 100644
--- a/swh/indexer/metadata_dictionary/cff.py
+++ b/swh/indexer/metadata_dictionary/cff.py
@@ -25,9 +25,9 @@ class CffMapping(YamlMapping, SingleFileIntrinsicMapping):
name = "cff"
filename = b"CITATION.cff"
mapping = CROSSWALK_TABLE["Citation File Format Core (CFF-Core) 1.0.2"]
- string_fields = ["keywords", "license", "abstract", "version", "doi"]
+ string_fields = ["title", "keywords", "license", "abstract", "version", "doi"]
date_fields = ["date-released"]
- uri_fields = ["repository-code"]
+ uri_fields = ["url", "repository-code"]
def _translate_author(self, graph: Graph, author: dict) -> rdflib.term.Node:
node: rdflib.term.Node
diff --git a/swh/indexer/tests/metadata_dictionary/test_cff.py b/swh/indexer/tests/metadata_dictionary/test_cff.py
index fb50ba5d..35964faf 100644
--- a/swh/indexer/tests/metadata_dictionary/test_cff.py
+++ b/swh/indexer/tests/metadata_dictionary/test_cff.py
@@ -85,6 +85,7 @@ RIS, schema.org, CodeMeta, and .zenodo.json.""",
"identifier": "https://doi.org/10.5281/zenodo.1162057",
"license": "https://spdx.org/licenses/Apache-2.0",
"version": "1.4.0-alpha0",
+ "name": "cffconvert",
}
assert expected == result
diff --git a/swh/indexer/tests/test_bibtex.py b/swh/indexer/tests/test_bibtex.py
index e0220d71..c2f4a861 100644
--- a/swh/indexer/tests/test_bibtex.py
+++ b/swh/indexer/tests/test_bibtex.py
@@ -7,7 +7,8 @@ import textwrap
import pytest
-from swh.indexer.bibtex import codemeta_to_bibtex
+from swh.indexer.bibtex import cff_to_bibtex, codemeta_to_bibtex
+from swh.model.swhids import QualifiedSWHID
def test_empty():
@@ -273,3 +274,101 @@ def test_affiliation():
}
"""
)
+
+
+def test_cff_empty():
+ assert cff_to_bibtex("") == textwrap.dedent(
+ """\
+ @software{REPLACEME
+ }
+ """
+ )
+
+
+def test_cff_invalid():
+ assert cff_to_bibtex("foo") == textwrap.dedent(
+ """\
+ @software{REPLACEME
+ }
+ """
+ )
+
+
+def test_cff_minimal():
+ assert (
+ cff_to_bibtex(
+ """
+cff-version: 1.2.0
+message: "If you use this software, please cite it as below."
+authors:
+ - family-names: Druskat
+ given-names: Stephan
+title: "My Research Software"
+date-released: 2021-08-11
+url: "http://example.org/"
+ """
+ )
+ == textwrap.dedent(
+ """\
+ @software{REPLACEME,
+ author = "Druskat, Stephan",
+ date = "2021-08-11",
+ year = "2021",
+ month = "08",
+ title = "My Research Software",
+ url = "http://example.org/"
+ }
+ """
+ )
+ )
+
+
+def test_swhid_type_snp():
+ assert codemeta_to_bibtex(
+ {
+ "@context": "https://doi.org/10.5063/schema/codemeta-2.0",
+ },
+ QualifiedSWHID.from_string(
+ "swh:1:snp:da39a3ee5e6b4b0d3255bfef95601890afd80709"
+ ),
+ ) == textwrap.dedent(
+ """\
+ @software{REPLACEME,
+ swhid = "swh:1:snp:da39a3ee5e6b4b0d3255bfef95601890afd80709"
+ }
+ """
+ )
+
+
+def test_swhid_type_rev():
+ assert codemeta_to_bibtex(
+ {
+ "@context": "https://doi.org/10.5063/schema/codemeta-2.0",
+ },
+ QualifiedSWHID.from_string(
+ "swh:1:rev:5b909292bcfe6099d726c0b5194165c72f93b767"
+ ),
+ ) == textwrap.dedent(
+ """\
+ @softwareversion{REPLACEME,
+ swhid = "swh:1:rev:5b909292bcfe6099d726c0b5194165c72f93b767"
+ }
+ """
+ )
+
+
+def test_swhid_type_cnt():
+ assert codemeta_to_bibtex(
+ {
+ "@context": "https://doi.org/10.5063/schema/codemeta-2.0",
+ },
+ QualifiedSWHID.from_string(
+ "swh:1:cnt:5b909292bcfe6099d726c0b5194165c72f93b767;lines=5-10"
+ ),
+ ) == textwrap.dedent(
+ """\
+ @codefragment{REPLACEME,
+ swhid = "swh:1:cnt:5b909292bcfe6099d726c0b5194165c72f93b767;lines=5-10"
+ }
+ """
+ )
--
GitLab